| Descriptive Statistics Table | ||||
|---|---|---|---|---|
| Mean, Min, Max, and Standard Deviation of Key Variables | ||||
| Variable | Mean | Min | Max | Standard Deviation |
| Official Development Aid (log) | 19.08 | 9.21 | 24.04 | 1.63 |
| Liberal Democracy Index | 0.25 | 0.00 | 0.86 | 0.20 |
| Overall Political Corruption Index | 0.57 | 0.01 | 0.96 | 0.25 |
| Executive Corruption Index | 0.57 | 0.01 | 0.98 | 0.26 |
| Public Sector Corruption Index | 0.54 | 0.00 | 0.98 | 0.26 |
| GDP per Capita (log) | 6.97 | 3.13 | 10.67 | 1.36 |
| Tax Revenue (% of GDP) | 16.00 | 0.00 | 147.64 | 8.16 |
| Trade (% of GDP) | 69.26 | 0.02 | 410.94 | 42.72 |
| Inflation (GDP deflator, %) | 39.46 | −100.00 | 15,444.42 | 391.59 |
| Exports (% of GDP) | 30.94 | 0.01 | 213.95 | 22.21 |
| Debt (% of GDP) | 51.83 | 2.99 | 277.53 | 31.47 |
| Labor Force with Basic Education (%) | 53.82 | 12.62 | 100.00 | 15.70 |
PSCI 3200 Final Project
Introduction
The recent slashing of the United States Agency for International Development (USAID) by the Trump administration has re-sparked the debate about whether countries should be responsible for distributing and providing foreign aid. Although this question about whether countries like the United States have a duty to provide aid is more of a philosophical question regarding global justice, this discussion can also be extended to the effectiveness of foreign aid. Foreign aid plays a critical role in shaping political and economic development with the intention of supporting state-building, promoting good governance, and reducing poverty across the Global South, but the actual impact of foreign aid remains ambiguous. A body of literature focuses on the effects that foreign aid has on corruption, where scholars argue that aid can promote transparency and reduce corruption by providing resources and incentives for institutional reform, including Burnside and Dollar (2000) and Okada and Samreth (2012). Others argue that aid fosters dependency, facilitates rent-seeking, and undermines domestic accountability, including Br"autigam and Knack (2004) and Knack (2001). However, the effects of foreign aid are largely believed to be influenced by institutional and contextual factors. My main research question concerns foreign aid’s impacts on corruption and using institutional quality as a moderator. Specifically, how does foreign aid impact corruption in developing countries, and how does regime type, specifically the level of democracy in a country, moderate this effect?
An early study by Burnside and Dollar (2000) found that foreign aid has a positive impact on growth in developing countries with good fiscal, monetary, and trade policies but has little effect in the presence of poor policies. A study by Okada and Samreth (2012) also found that foreign aid generally reduces corruption, and its reduction effect is greater in less corrupt countries. This context-based literature also extends to economic complexity, where foreign aid reduces economic complexity in countries with lower initial levels, but has a positive effect in countries with higher economic complexity according to Kamguia et al. (2022). Addionally, Rahnama, Fawaz, and Gittings (2017) finds that foreign aid increases economic growth in high-income developing countries but reduces growth in low-income developing countries. Other research focuses on the effects that foreign aid has on state capacity. For example, Br"autigam and Knack (2004) found that higher aid dependence is associated with weaker bureaucratic quality and greater corruption, particularly in Sub-Saharan Africa. Similarly, Morrison (2007) and Knack (2009) show that states dependent on foreign aid or natural resources have lower-quality governance due to weakened incentives for accountability and reform. In other words, recent studies find that the effects of foreign aid can have negative impacts on state capacity and governance. Andersen, Johannesen, and Rijkers (2022) also used data on offshore banking activity and found that disbursements of foreign aid are associated with increased deposits in offshore accounts controlled by elites in the recipient countries. This pattern is especially pronounced in non-democratic and high-corruption contexts, offering robust evidence of elite capture of aid flows. To address these risks, some donors use bypass aid, routing funds through NGOs or multilateral agencies rather than recipient governments. Dietrich (2013) shows that donors strategically bypass state institutions in countries with low governance quality or high corruption levels, showing that regime type moderates the relationship between aid and governance outcomes, not just in terms of effectiveness, but also in terms of how aid is delivered. Despite the rich literature on foreign aid and corruption, few studies have directly tested how democracy moderates the relationship between foreign aid and corruption. I hypothesize that foreign aid reduces corruption in democracies, where institutional constraints and accountability mechanisms are stronger, but increases corruption in autocracies, where aid is more easily captured by elites.
Theory and Hypothesis
This project investigates how the effect of foreign aid on corruption varies depending on a country’s level of democracy. Specifically, I examine whether democratic institutions mitigate or exacerbate the risk that foreign aid contributes to corrupt practices. My central hypothesis is that foreign aid reduces corruption in countries with high democracy scores, while it increases corruption in countries with low democracy scores. In other words, the relationship between foreign aid and corruption is conditional on the strength of democratic institutions and the quality of governance. Hypothesis: Foreign aid reduces corruption in countries with high democracy scores but increases corruption in countries with low democracy scores.
This hypothesis is grounded in a growing body of literature suggesting that the effectiveness of foreign aid is highly context-dependent. Scholars, like the ones I have mentioned above, have found that aid does not have monolithic effects, but instead interacts with domestic political institutions in complex ways. For the sake of clarity, when I refer to a democratic country, I am really referring to countries that have characteristics of a liberal democracy, defined as “a form of political system where enfranchised citizens elect a government of representatives, combining democratic principles with liberalism” with “civil and political rights as well as individual freedoms, aiming to limit the powers of the state and prevent a tyranny of the majority” “Liberal Democracy” (n.d.).
Countries with strong democratic institutions, theoretically, have greater institutions that ensure transparency and accountability, including elections, independent media, and the rule of law. These characteristics, I predict, should allow foreign aid to be distributed as intended. In other words, aid is more likely to pass through official channels and subjected to scrutiny, decreasing the opportunity for embezzlement and misallocation. On the other hand, autocratic countries lack these basic democratic structures, creating contexts where dissent and opposition are often suppressed. As a result, I expect that foreign aid may be captured by elites or diverted for political or personal gain. This makes aid more likely to subject to misuse by leaders, creating a principal-agent problem concerning the allocation of aid. Thus, I expect a negative effect on corruption from an interaction between foreign aid and democracy because democracies, theoretically, have greater institutions that ensure transparency and accountability, leading to greater allocation of resources and less corrupt practices. However, countries that have less democratic institutions, including autocracies, are expected to have weaker mechanisms that monitor the allocation of foreign aid and resources in general.
This logic is consistent with previous empirical findings in the literature, where stronger institutions usually have a greater effect on foreign aid success and economic growth. Scholars like Okada and Samreth (2012) have shown that the impact of foreign aid on corruption is dependent on the recipient country’s existing levels of corruption and institutional quality. They find that foreign aid can reduce corruption, but mainly in countries where institutions are already strong and minimal existing corruption. I expect my findings to have similar outcomes. Similarly, other studies that I have provided suggest that aid can promote good governance and reform when the political environment is already in favorable conditions. In other words, the impact that foreign aid has on corruption can be thought of as a positive feedback loop, where process where good institutional conditions create a pathway for stronger institutions. Thus, democratic institutions, as opposed to authoritarian ones, will have a greater impact on reducing corruption, whereas non-democratic regimes will be bolstered by foreign aid, increasing corruption.
Research Design
The research design for this analysis includes data from the World Bank Group, World Development Indicators (WDI), and V-Dem. To measure foreign aid, I am using Official Development Assistance (ODA) data from the Organization for Economic Co-operation and Development (OECD) and the World Bank Group. ODA refers to “government aid that promotes and specifically targets the economic development and welfare of developing countries.” I incorporate ODA data specifically because ODA flows comprise contributions of donor government agencies to developing countries, making it easy to determine the levels of aid that each developing country is receiving. It is also important to note that ODA receipts comprise disbursements by bilateral donors and multilateral institutions, meaning ODA flows are considered to already have been delivered rather than “promised.”
To measure democracy and corruption, I am using variables from the V-Dem Institute, a database that measures the qualities of different democracies. In my analysis, I will be using “v2x_libdem,” or the liberal democracy index, to measure democratic institutions in a country. This variable aims to measure the quality of democracy by the limits placed on government. This is a useful way to measure democracy because it follows my reasoning above, where countries with these types of limited institutions should create an environment for foreign aid to reduce corrupt activities.
I will also be using “v2x_corr,” or the political corruption index, to measure a country’s level of corruption. This variable measures the extent of both petty and grand corruption, including bribery and embezzlement, and aggregates corruption among top elites and the broader public sector into one index. My empirical extension will include “v2x_execorr” and “v2x_pubcorr” to distinguish between executive corruption and public sector corruption, respectively.
My control variables are from the World Development Indicators (WDI). The WDIs were created by the World Bank to track progress on global development and the fight against poverty, including variables on economic and social development. I include GDP per Capita, Tax Revenue, Trade Openness, Debt Level, and Labor Education as covariates because they have the potential to influence foreign aid and corruption. GDP per capita is included because wealthier countries tend to have stronger institutions and lower corruption, and may receive different levels of aid. Tax revenue (% of GDP) reflects state capacity and fiscal accountability, which influence both aid dependency and corruption. Trade openness (% of GDP) captures economic integration, which can either reduce corruption through transparency or increase it through illicit flows. Debt (% of GDP) indicates fiscal stress and governance quality, affecting both aid allocation and corruption risks. Labor force education proxies for human capital and civic engagement, which can be linked to less corruption and potentially higher aid capacity.
Since my hypothesis implies a causal relationship, that foreign aid affects corruption differently depending on regime type, there are several threats to interpreting the coefficient on the aid-democracy interaction as a causal effect. First is reverse-causality, where corruption may influence where foreign aid goes, such as the example of bypass aid. I address this by lagging my independent variables, foreign aid and democracy, by 5 years. I do not expect the effects of foreign aid to impact corruption immediately, and this lag will address the issue of reverse-causality by making it clear that foreign aid should impact corruption, not the other way around. Another threat is omitted variable bias, where I cannot include all potential variables like path dependency theory of colonial institutions or other aspects of institutional quality, like foreign influence. Although I attempt to mitigate these threats by including key control variables and country and year fixed effects, establishing a clean causal effect remains challenging.
Although the empirical test that I have proposed in my research design directly measures the relationship of foreign aid and democracy with corruption, it does not create a distinction between the types of corruption that foreign aid can influence. I have already mentioned that the variable I use creates an index for both petty corruption, or everyday abuse of power by public officials in their interactions with ordinary citizens, and grand corruption, or abuse of high-level power by public officials that benefits a select few at the expense of the many. My empirical extension will focus on whether there is a distinct impact that foreign aid and democracy have on petty versus grand corruption. For example, one might expect that in low-democracy countries, foreign aid could exacerbate grand corruption by empowering elites who control state resources, whereas in high-democracy countries, stronger institutions might prevent elite capture but still struggle with petty bureaucratic inefficiencies. This extension contributes to the broader literature by offering an observable implication that helps probe the mechanisms underlying the aid-corruption relationship. If foreign aid is more strongly associated with grand corruption in autocratic contexts but not in democratic ones, it would provide more specific evidence of elite capture as a pathway through which aid influences governance outcomes. Conversely, if the pattern is more pronounced with petty corruption, it may indicate that bureaucratic capacity, rather than political incentives, plays a larger role.
Below is the full model equation as well as the descriptive statistics of the final data set used for the following analysis.
\[ Corruption_{it} = \beta_0 + \beta_1 Aid_{i,t-5} + \beta_2 Democracy_{i,t-5} + \beta_3 (Aid_{i,t-5} \times Democracy_{i,t-5}) + \gamma_1 \text{GDP per Capita}_{it} + \gamma_2 \text{Tax Revenue}_{it} + \gamma_3 \text{Trade}_{it} + \gamma_4 \text{Inflation}_{it} + \gamma_5 \text{Debt}_{it} + \gamma_6 \text{Labor Education}_{it} + \varepsilon_{it} \]
Results
The hypothesis that foreign aid reduces corruption in countries with high democracy scores but increases corruption in countries with low democracy scores is supported by the following analysis. Figure 1 below shows the raw relationship between foreign aid and corruption using democracy as a grouping variable. I differentiate between Low, Medium, and High levels of democracy by simply grouping the corruption variable into thirds, where the bottom third represents low levels of democracy, the middle third represents medium levels of democracy, and the upper third represents high levels of democracy. This graph shows the basic relationship between foreign aid and corruption at different levels of democracy, where foreign aid to countries with high levels of democracy experience corruption at a less extreme effect, whereas foreign aid to countries with low levels of democracy experience more levels of country.
Table 2 below confirms this same result through regression analysis. Without control variables, a one-unit increase in log ODA, while holding democracy levels at zero, is associated with a small, yet statistically significant, increase in corruption, suggesting that aid tends to fuel corruption in the absence of strong institutions. An interaction effect between ODA and democracy levels also suggests that there is a small decrease in corruption for countries with higher levels of democracies. It is also interesting to note that more democratic countries, holding foreign aid at zero, experience less corruption, although this effect is statistically insignificant. Incorporating control variables strengthens these results, where a one-unit increase in log ODA, while holding democracy at zero, is associated with a 0.040 percentage point increase in corruption. This effect reverses in democratic countries, foreign aid is associated with a -0.44 decrease in corruption. These results are also statistically significant with a p-value less than 0.05. In other words, foreign aid is associated with increased corruption in undemocratic countries, while decreased in democratic countries, supporting my hypothesis. It is interesting to note that the regression also suggests that countries with higher democracy levels, holding foreign aid at zero, is associated with a 1.688 percentage point increase in corruption. However, this effect decreases, as shown by the interaction, once foreign aid is introduced, suggesting that foreign aid can minimize corruption in democracies. In other words, this model appears to show that democratic countries are more corrupt, but this does not align with the literature because democratic countries are believed to address corruption better. This is an interesting phenomena to explore, but beyond the scope of this paper. Figure 2: Marginal Effects Plot of ODA and Corruption by Democracy shows this effect visually below, where, although democratic countries are shown to have higher levels of corruption, foreign aid decreases corruption. This opposite effect is true for less-democratic countries, where these countries have lower levels of initial corruption, but foreign aid increases levels of corruption.
| No Controls | With Controls | |
|---|---|---|
| Intercept | 0.408*** | -0.033 |
| (0.037) | (0.381) | |
| ODA (log) | 0.011*** | 0.040* |
| (0.002) | (0.017) | |
| Democracy | -0.185 | 1.688* |
| (0.120) | (0.668) | |
| ODA × Democracy | -0.003 | -0.084* |
| (0.006) | (0.034) | |
| GDP per Capita (log) | 0.016 | |
| (0.032) | ||
| Tax Revenue (% GDP) | 0.002 | |
| (0.003) | ||
| Trade (% GDP) | -0.000 | |
| (0.001) | ||
| Inflation (GDP deflator, %) | 0.000 | |
| (0.001) | ||
| Exports (% GDP) | 0.001 | |
| (0.002) | ||
| Debt (% GDP) | 0.001 | |
| (0.000) | ||
| Labor Education (%) | -0.002** | |
| (0.001) | ||
| + p < 0.1, * p < 0.05, ** p < 0.01, *** p < 0.001 | ||
Empirical Extension
The analysis I have conducted supports my hypothesis, but I now wish to determine whether foreign aid has different effects on different types of corruption, including grand corruption and petty corruption. A simple look at the raw effects of ODA on Grand Corruption versus Petty Corruption in Figures 3 and 4 shows a similar relationship between grand corruption and petty corruption across different levels of democracy.
A comparison between Table 3 for grand corruption and Table 4 for petty corruption confirms a similar relationship between the two, but there appears to be a stronger magnitude of effect of foreign aid on grand corruption. For example, a one-unit increase in log ODA, holding democracy levels fixed at zero, is associated with a 0.067 percentage point increase in grand corruption, but only a 0.033 percentage point increase in petty corruption. Similarly, adding an interaction between log ODA and democracy levels reverses the effect for democratic countries, where grand corruption levels decrease by -0.066 percentage points and petty corruption levels decrease by -.040 percentage points. In other words, foreign aid has a greater magnitude of an effect on grand corruption than petty corruption. Like the above analysis, it is interesting to observe the respective levels of corruption in democracies in each of these regressions. For grand corruption, democratic countries appear to have a 2.603 percentage point increase in corruption without foreign aid. For petty corruption, democratic countries appear as less corrupt with a 1.507 percentage point increase in corruption. Figures 5 and 6 visually show these results and do not deviate from my initial findings. Thus, this empirical extension can rule out the fact that foreign aid impacts grand corruption and petty corruption differently because these results demonstrate a similar impact on the two different types of corruption.
| No Controls | With Controls | |
|---|---|---|
| Intercept | 0.465*** | -0.150 |
| (0.047) | (0.524) | |
| ODA (log) | 0.008*** | 0.067** |
| (0.002) | (0.023) | |
| Democracy | -0.267+ | 2.603** |
| (0.151) | (0.920) | |
| ODA × Democracy | -0.000 | -0.133** |
| (0.008) | (0.046) | |
| GDP per Capita (log) | -0.011 | |
| (0.044) | ||
| Tax Revenue (% GDP) | 0.002 | |
| (0.005) | ||
| Trade (% GDP) | 0.000 | |
| (0.001) | ||
| Inflation (GDP deflator, %) | -0.001 | |
| (0.001) | ||
| Exports (% GDP) | -0.001 | |
| (0.003) | ||
| Debt (% GDP) | 0.001 | |
| (0.001) | ||
| Labor Education (%) | -0.004*** | |
| (0.001) | ||
| + p < 0.1, * p < 0.05, ** p < 0.01, *** p < 0.001 | ||
| No Controls | With Controls | |
|---|---|---|
| Intercept | 0.257*** | -0.479 |
| (0.045) | (0.352) | |
| ODA (log) | 0.018*** | 0.033* |
| (0.002) | (0.015) | |
| Democracy | -0.027 | 1.507* |
| (0.145) | (0.618) | |
| ODA × Democracy | -0.012 | -0.073* |
| (0.007) | (0.031) | |
| GDP per Capita (log) | 0.057+ | |
| (0.029) | ||
| Tax Revenue (% GDP) | 0.006+ | |
| (0.003) | ||
| Trade (% GDP) | -0.001 | |
| (0.001) | ||
| Inflation (GDP deflator, %) | -0.000 | |
| (0.001) | ||
| Exports (% GDP) | 0.002 | |
| (0.002) | ||
| Debt (% GDP) | 0.001** | |
| (0.000) | ||
| Labor Education (%) | -0.002* | |
| (0.001) | ||
| + p < 0.1, * p < 0.05, ** p < 0.01, *** p < 0.001 | ||
Discussion and Policy Implications
This analysis offers clear evidence that the effectiveness of foreign aid in reducing corruption is strongly conditioned by a country’s level of democracy. Specifically, the analysis shows that foreign aid is associated with increased corruption in autocratic countries but reduced corruption in democratic ones, supporting my initial hypothesis. The results are statistically significant and hold across different levels of corruption, including grand and petty corruption. The interaction between foreign aid and democracy is statistically significant and substantively meaningful, suggesting that institutional quality plays an important role in shaping the outcomes of aid.
These results support prior literature based on conditionality, where aid is not inherently good or bad, but its impact depends on the context of the recipient country. Although the regression models show that democratic countries have higher initial levels of corruption, which is inconsistent with my assumption that democratic countries have lower levels of corruption, the interaction term shows that foreign aid has varying outcomes on corruption depending on levels of democracy. In other words, foreign aid decreases corruption in countries with higher levels of democracy, whereas foreign aid increases corruption in countries with lower levels of democracy. The empirical extension further confirms that these patterns apply to grand and petty corruption, although the effects appear stronger for grand corruption.
The policy implications are that aid allocation strategies should account for recipient regime type. Donor countries should consider bypass aid when considering foreign aid for autocratic countries, largely due to the negative effects that it may have on grand and petty corruption. Additionally, donor countries should continue to make aid conditional based on democratic values, where recipient countries should already have institutions in place that exemplify democratic ideals and processes. They should also pair aid with investments in democracy promotion and institutional reform, which appear to make aid more effective at reducing corruption.
Although the results are interesting, this analysis does have limitations. Establishing causality is difficult, and omitted variable bias or measurement error could still influence these results, including the influence of aid that is already conditional and the interfering effects of such bureaucratic processes. Similarly, the findings concerning initial levels of corruption should be explored further because they contradict existing literature about the role of democracy in limiting corruption. Future research should attempt to address these issues, particularly through new data and methods.
References
Andersen, Jørgen Juel, Niels Johannesen, and Bob Rijkers. 2022. “Elite Capture of Foreign Aid: Evidence from Offshore Bank Accounts.” Journal of Political Economy 130 (9): 2360–2404.
Br"autigam, Deborah A, and Stephen Knack. 2004. “Foreign Aid, Institutions, and Governance in Sub‐Saharan Africa.” Economic Development and Cultural Change 52 (2): 255–85.
Burnside, Craig, and David Dollar. 2000. “Aid, Policies, and Growth.” American Economic Review 90 (4): 847–68.
Dietrich, Simone. 2013. “Bypassing Government: Foreign Aid, NGOs, and Democratic Accountability.” International Studies Quarterly 57 (4): 698–712.
Kamguia, Brice, Sosson Tadadjeu, Clovis Miamo, and Henri Njangang. 2022. “Does Foreign Aid Impede Economic Complexity in Developing Countries?” International Economics 169 (May): 71–88. https://doi.org/10.1016/j.inteco.2021.10.004.
Knack, Stephen. 2001. “Aid Dependence and the Quality of Governance: Cross-Country Empirical Tests.”
———. 2009. “Sovereign Rents and the Quality of Tax Policy and Administration.” Journal of Comparative Economics 37 (3): 359–71.
Morrison, Kevin M. 2007. “Natural Resources, Aid, and Democratization: A Best-Case Scenario.” Public Choice 131 (3-4): 365–86.
Okada, Keisuke, and Sovannroeun Samreth. 2012. “The Effects of Foreign Aid on Corruption: A Quantile Regression Approach.” Economics Letters 115 (2): 240–43.
Rahnama, Masha, Fadi Fawaz, and Kaj Gittings. 2017. “The Effects of Foreign Aid on Economic Growth in Developing Countries.” The Journal of Developing Areas 51 (3): 153–71. https://doi.org/10.1353/jda.2017.0066.